A Bull Market In Bad Predictions

Includes: DIA, QQQ, SPY
by: Jeff Miller

Why does this happen whenever I try to take a few days off? The market for dubious predictions has geared up in earnest!

While on vacation I was watching the market (but without my customary TIVO), events developed exactly as I predicted. I warned about signal and noise, the challenge to traders, and the opportunity for long-term investors.

I have also been reading Nate Silver's book, The Signal and the Noise, which includes a lot of wisdom on these topics. I plan a full review when I finish. One of Silver's points concerns predictions without any confidence interval. Many themes will be familiar to readers of "A Dash" since I highlight pundits who claim expertise outside of their "happy zone." Let us highlight the three worst items from the past week.

  • The fiction -- the ECRI claims that we are now in a recession. This is ECRI 4.0 after their 2011 forecast failed, their revised 2012 forecast failed, and their complaint about seasonal adjustments being wrong has not proven out. They are now playing out the last straw, that they are the only ones who can forecast recessions in advance and that no one else knows until after it is over. This will obviously require a deeper look. Let me cite the most obvious incorrect statement in their claims: The business cycle has peaked and they are the only ones who know this.

The reality. No one knows whether the current period will eventually be defined as a recession. A recession requires a significant decline (which you do not know until you have seen it). At that point the NBER goes back to the last peak. The ECRI presentation last week "assumed facts not in evidence." They are ignoring the reduction in business spending before the election and the fiscal cliff. They are exploiting the Super storm Sandy effects. We can expect them to pound the drum even more during the next month, since the weak patch will take a couple of months to sort out.

I have a personal sadness about this, since I like and admire the ECRI principals. I am going to write another piece about how and why their methods failed. I wish that they had just been willing to accept the changing evidence -- and maybe open the kimono a little bit.

  • The fiction --- the decline to zero growth. GMO's Jeremy Grantham opines that the US economy is on a zero growth path until 2050. He focuses on the two best drivers of growth -- population and productivity. In this CNBC segment Maria Baritromo breathlessly praises Grantham:

...He gets paid to make predictions, Steve. that's what he's doing. by the way, his former predictions have been right. let's give him that.

The reality. No one knows what will happen in 2050. Grantham has ignored a decline in immigration (something that has helped US GDP in the past) to support his perma-bear position. Pretending to this kind of knowledge gets headlines, but should be a warning signal to investors. The media commentary points out that he manages a gazillion dollars or so. Maybe a few of those investors should look to managers who are more grounded in facts.

And by the way, maybe Maria should cite Grantham's track record -- 47% -- before claiming that he has made so many great calls. Anyone who takes this silly prediction seriously should look back forty years for a comparison.

  • The fiction. The latest new and greatest recession indicator. This is from Lance Roberts (who without apology highlighted the bogus 100% recession indicator). He is now back with a new entry, endorsed by John Hussman. Roberts takes some existing economic forecasting indicators that do not initially give the result he hopes for. He then does some arithmetic and creates something that has a lame correlation to past recessions. Hussman (who does similar things) embraces this approach.

The Reality. The St. Louis Fed creates about 60,000 data series. If you do some math transformations as Roberts did, you can turn this into a million or so possibilities. If you then set a "trigger" at an arbitrary level based upon a handful of past cases (the way Hussman does) you can multiply this into the hundreds of millions range.

It is bad research, bad methodology, and a seriously misleading result. It is impossible to prove, since the bad guys used all of the data. There is nothing left to prove them wrong.


So much bogus commentary, and so little time. Can't a guy take a few days off?

I will follow up on all of these themes. Here are the main ideas:


The ECRI errors will require a more careful review -- it is on my agenda. Meanwhile, you can get the basic concept of their mistake by reviewing my recession forecasting page.

Fiscal Cliff

This theme continues with silly trading in the absence of information. Listen up! We have no new information since the election. We will not know anything new for a few weeks. Trade at your peril.


It almost seems too obvious. So many have so much at stake in scaring investors. They have clearly won the battle, with aggressive money flowing into anything with a high yield and conservative money going to farmland and ammunition. My conversations with investors show that many are scared witless (TM OldProf).

Since the big rewards go to the contrarian investor, there are some great opportunities.

I like CAT as the proxy stock for an economic rebound, although it is (incorrectly) China-centric. I also like some health care insurers and defense stocks -- UNH and LMT as examples -- as winners in the fiscal cliff compromise. AFL is a good play if there is no disaster in Europe.

These are complex questions, so I plan to write more on each issue.

Disclosure: I am long CAT, AFL, UNH, LMT. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it. I have no business relationship with any company whose stock is mentioned in this article.